Alcohol use is a major contributor to the leading causes of adolescent morbidity and mortality and to other health risks, including other substance use and risky behaviors. Adolescent alcohol use may also result in long term cognitive, behavioral, anatomic and neurochemical changes. Additionally, initiating alcohol use in adolescence significantly increases a person's risk for adult alcohol use disorders. Given this considerable public health burden, successfully identifying adolescents at risk for current or future problematic alcohol use is of major importance, as intervention may significantly affect both adolescent and adult alcohol use disorders. The National Institute of Alcohol Abuse and Alcoholism (NIAAA) has developed a novel two-question screening tool for the early detection of alcohol use and problems. Empirically based, the screen may be a powerful predictor of current and future negative consequences of alcohol use. We propose to test the goals of RFA- AA-12-008 by validating the NIAAA two-question screen with 12-17 year olds in 15 U.S. pediatric emergency departments (PED), by pursuing two specific aims: Primary Aim 1) To determine the concurrent, convergent, discriminant, and predictive validity of the two-question screen regarding alcohol misuse, and Secondary Aim 1) To determine the concurrent and predictive validity of the two-question screen regarding other substance use and risky behaviors. The advantage of the PED setting is the ability to capture high risk populations such as school dropouts and those without primary care or insurance, who often use PEDs for their medical care and are missed in other study settings. Employing the Pediatric Emergency Care Applied Research Network (PECARN), we will use a multi-center design to test the concurrent, convergent and predictive validity of the NIAAA two-question screen using established, validated measurement instruments in a very large, diverse sample of youths. Youths 12-17 years (N=8000) presenting to participating PECARN PEDs will be enrolled over a 3.5-year period and will complete the two-question screen and an alcohol and other drug (AOD) and behavior (criterion) assessment battery. Responses from these youths will be used to assess the concurrent convergent, and discriminant validity of the NIAAA two-question screen. Two subsamples of youths will be derived from the initial sample. The first subsample (N=300) will be re-administered the NIAAA two-question screen, one week after their initial ED visit to assess the test-retest reliability of the NIAAA two-question screen. The second sample (N=2500), will be reassessed at 12 and 24 months after their PED visit with the criterion assessment battery to assess the predictive validity of the two-question screen. This large sample will permit stratified analyses by gender, race, ethnicity and other important demographic variables, which are not possible with single center or smaller sample studies. The proposed research is significant because it rigorously evaluates the ability of the NIAAA two-question screen to predict current and future AOD risks as well as other behavior problems, and will generate very robust, generalizable results.